Chương 66: Last day

He had the data. He had had the data since approximately 1420 the previous afternoon, which meant he had been sitting on a completed assessment for eleven hours, and the assessment was not why he was still on the ridge, and he had run out of road on not examining that. Day three began at 0600 when the group below broke camp. They were faster than Day 1. Measurably: he tracked the sequence from fire-down to movement-ready and the elapsed time was seventeen minutes shorter than the first morning he'd watched them.

The improvement was distributed across all three rather than concentrated in one. Marco's pack consolidation more efficient, Dez's gear check without the two false-starts he'd run the first morning, Yuki's treatment kit stowed in a sequence that suggested she'd reorganized it at some point in the last two days for faster access. Seventeen minutes. Compounded across the days they'd been together before he found them, across the days ahead of Haven's Reach, the number would keep dropping.

Groups improved or they collapsed, and the behavioral signatures he'd been reading for three days pointed unambiguously in one direction. He had been an indirect cause of some of the improvement. He was aware of this. He had been adjusting his patrol pressure, pulling back from the section of the corridor south of their camp on Day 2, which had reduced the threat density they were moving through and allowed them to focus attention on efficiency rather than engagement.

He had done this without deciding to, which was the kind of thing he filed in the unnamed column alongside everything else that didn't have a standard header. I have been running covert threat management for three people I haven't spoken to for three days. From a ridge. This is completely normal behavior and I will not be taking questions.

Day three of the program, the narration began, and there was something different about it this morning, something that had less energy behind it, the way a joke starts to change shape when the thing underneath it gets too close to the surface to stay funny. The analyst remains on his ridge. The subjects have improved their camp-break efficiency by seventeen minutes, which the analyst has catalogued with the precision of a man who has mistaken data collection for participation.

The production team would like to note that the analyst has also been clearing threat signatures from the group's southern approach corridor without conscious authorization from the part of his brain that is supposedly running this operation. He is helping them. He has been trying to help them since Day 1. He has filed this under "unnamed" because naming it would require admitting it is happening, and admitting it is happening would require descending from the ridge, and descending from the ridge would require...

The narration stopped. Not because he'd ended it. Because it had arrived at the same wall the rest of him had been sitting against for eleven hours, and the documentary format didn't have a way through it either. 【 AETHERFORGE NEXUS 】 Assessment — Complete PK probability: 9% (negligible) Behavioral classification: Genuine survivors, non-predatory Confidence: 91% — threshold exceeded by 6 points The Watchers note 91% has been sitting on your bark-sheet since yesterday afternoon. Six points above threshold.

Eleven hours above threshold. The number is not the problem and you know the number is not the problem, which means you also know what the problem is, which means the only remaining question is how long you plan to stay on this ridge calling it data collection. He read it and sat with it. The Watchers had a word for the gap. So did he. It was fear, which he was allowed to have, dressed up in methodology.

Not fear of Marco, Yuki, and Dez specifically, nine percent was negligible, he cleared worse odds before breakfast most days. Fear of the category. Of what happened to your internal threat-mapping when people stopped being variables and started being something your brain tracked without asking your permission first, the way his had been tracking these three from a ridge for eleven hours while he called it research. He looked at them moving north.

Marco out front, covering ground too fast the way he always did before his rhythm settled. Dez already asking something. Yuki somewhere between them, moving the way she moved. The thing about fear dressed as methodology was that it was still fear. He knew what the ridge had been. He knew what the blank column header was. He knew the Watchers knew, which was presumably why they were enjoying themselves. Tomorrow morning. He'd have something useful to say by then.

He hoped they'd still be there to say it to, which was its own kind of answer about what the last three days had actually been. The decision, once made, had the specific texture of decisions that had been available for longer than he'd admitted. Not a revelation.

More like: he had been standing in a doorway for eleven hours and had now taken a step through it, and the step was not dramatic, it was just a step, and the other side of the doorway was terrain he didn't have good data on and would have to map as he moved through it. He knew how to map unfamiliar terrain. He had been doing nothing else for thirteen days. The author's tale has been misappropriated; report any instances of this story on Amazon. The first sentence was the problem.

He had approximately fourteen hours and he was going to use some portion of them on the problem of the first sentence, which was not the most efficient allocation of fourteen hours but was the one he was committed to. He had been alone for a hundred and twenty-nine days.

The last thing he had said to another human being, the last real exchange, not a threat response, not a channel post, had been a performance review follow-up email in Sacramento that he had written at 4: 47 PM on a Tuesday and which had contained the phrase please advise on timeline three times. He was not going to open with please advise on timeline.

He thought about it while the group moved north and he tracked at distance, running the behavioral read on autopilot while the front of his brain worked the sentence problem. The first sentence needed to be useful. It needed to be true.

It needed to not be hello because hello was a social convention from a context that no longer applied and also, if he was being honest, he did not know if he could produce hello with a straight face after a hundred and twenty-nine days of the Crimson Wilds redefining what a greeting was for. Something tactical. Something accurate. Something that opened a conversation rather than closed one. 【 AETHERFORGE NEXUS 】 Willpower +0. 2 Decision executed: 130-day isolation pattern overridden.

Category: social re-engagement, voluntary. The Watchers note this is the largest single WIL expenditure you have made without a combat trigger. For reference, the list of things that have cost you less Willpower includes: two near-death meter events, one permadeath scare, and seventeen days of sustained starvation. You have decided to approach three people you have been watching for three days. The Watchers find your threat-assessment system very thorough.

They find this particular threat assessment the most thorough of all. He read it twice. Largest single WIL expenditure without a combat trigger. He had spent Willpower on sustained adaptation, on conditions that actively tried to kill him, on the specific grinding cost of staying functional when the environment wanted him not to be. The Nexus had logged all of it and this was still the largest single draw. He sat with that for a moment.

And willingly ignored it, hoping that this decision would also give him some sort of willpower boost. Then he went back to working the sentence problem, because that was a solvable problem, and what he'd just learned about himself was not something he needed to solve tonight. The Watchers could document it. He had a first sentence to finish. By 1400 he had four versions.

He ran them against his assessment of the group's likely response patterns, Yuki's data-seeking, Marco's directness, Dez's immediate follow-up questions, and evaluated each version for information density, opening value, and the likelihood of producing a response he could work with. Version one: threat warning. Accurate, immediately useful, no personal information required. Closed on follow-up. Too narrow. Version two: introduction first, context second. Sequencing was wrong.

Introduction without utility was a social convention and social conventions without utility were noise. Version three: route intelligence. Specific, high value, required too much explanation to be a first sentence. Version four arrived at 1600 and it was right. He knew it was right because it was tactical and true and opened toward something rather than closing, and it contained enough utility to justify the approach without requiring him to be more than he currently knew how to be in the direction of other people.

He wrote it on the bark-sheet. Then he crossed it out, because writing the first sentence you were going to say to another human being after a hundred and thirty days felt like something that required a category of sincerity he was not ready to commit to paper. He kept it in his head instead. Ninety-one percent. I have made harder calls on worse data many times. Tomorrow morning. I have approximately fourteen hours to figure out what my first sentence to another person in 130 days should be. Camp. Ridge.

Third night. He set the perimeter in the dark by feel and memory, which was not tactically optimal and which he permitted himself because the Fatigue meter was at seventy-four and he needed the time he would have spent on the perimeter assessment to sleep instead.

Below him the group's fire was the same size it had been every night, too large, warm, visible from the ridge in a way that was a security liability and also the way three people looked when they had decided that warmth was worth something, which was a calculation he had been making differently for four months and was possibly in the process of revising. He revised the first sentence twice more before sleep. The third version was worse than the fourth. The fourth was still correct. He kept the fourth.

This is the series finale, the narration said, and the voice was quiet, and it was not funny, and he let it talk anyway because it had earned the closing monologue by being the only thing in his head willing to say what was actually happening. Tomorrow morning the analyst will descend from the ridge.

He will walk toward three people he has been watching for seventy-two hours with the focused attention of a man who forgot what it felt like to be in the same room as another person and has been relearning it at a distance because the distance was the only version he could manage. He will say eleven words. He has tested the eleven words against three behavioral profiles and four alternative phrasings and has selected the version with the highest probability of producing a conversation rather than a silence.

He has done all of this because he is afraid, and because he is a data analyst, and because those two things have been the same thing for a hundred and thirty days. The production team would like to thank the Watchers for their ongoing viewership. The production team would like to thank the broadleaf cluster for its structural support.

The production team would like to thank the analyst for eventually, after seventy-two hours and ninety-one percent confidence and a Willpower expenditure that exceeded two near-death events, deciding to go say hello. Tomorrow: a different show. It was tactical. It was the best version he was capable of.

He closed his eyes and the fire below him burned too large and too warm in the dark, and tomorrow he was going to walk down the ridge and say eleven words to three people he had been watching for three days, and that was going to have to be sufficient because he had the data and the data said go and he had decided and the decision held. He slept.